As promised, today's post will be Part II of my Questions & Answers About Adverse Impact. In case you missed it on Tuesday, check out Part I here. When we left off, I was addressing whether or not it's legal to use a test that has adverse impact. So let's pick up from there:
What if there is no bottom line adverse impact, but there is adverse impact at different phases of the process?
This question is relevant to a multi-stage selection process. For example, an automated application process, followed by a comprehensive test battery and a structured interview, assuming only applicants who pass the previous stage would move to the next step in the selection process. If no adverse impact is found at the final stage, employers may mistakenly disregard the importance of a validation study, which is understandable. If there is no bottom line adverse impact, the Uniform Guidelines state that "in most circumstances there is no obligation under the Guidelines to investigate adverse impact for the components, or to validate the selection procedures used for that job." Of course, if adverse impact is found for the total selection process, the individual components should be evaluated to identify the underlying cause.
However, caution is warranted. The Supreme Court has proven willing to overturn the bottom line concept dictated in the Uniform Guidelines (Connecticut v. Teal, 1982). In this case, there was no bottom line adverse impact, but a disparity in pass rates at one stage sufficed to establish a prima facie case for adverse impact. The safest approach is to assess the existence of disparity in pass rates at the final stage, and at each step (whenever possible), documenting relevant validity evidence for each selection element.
What is the legal process for a discrimination claim involving a test with adverse impact?
Applicants or employees for private businesses with alleged discrimination cases have the right to sue individually or under the EEOC. Applicants or employees within the OFCCP's jurisdiction have no right to sue, but the OFCCP can impose remedies on contractors without court action. Most employment discrimination lawsuits are resolved outside of court. But understanding the judicial process during a trial helps employers prepare a legally defensible selection system in the first place.
Adverse impact cases involve 3 phases where the burden shifts between claimant and employers, and the trial can end in any phase:
- Phase I - Proving Adverse Impact
- Phase II - Employer Defense
- Phase III - Less Discriminatory Alternatives
The take-away here is that employers need to continually monitor their selection procedures for adverse impact and proactively consider viable alternatives if adverse impact is present.
How does the practice of matching employee representation with that of the local community work?
A number of organizations approach the issue, not from the components of the selection system per se, but from a comparison of the local labor marketing and their own distribution of minority groups. They try to mirror their workforce to match the percentages in the local labor market. A couple problems with this approach: The first is that the definition of the labor market is ambiguous. Second, it doesn't focus on quality of hires, but only on relevant percentages.
Rather than trying to match your employee representation to that of some defined applicant pool, Select International strongly recommends that you simply focus on developing job-relevant and valid selection practices to begin with, as a means of dealing with adverse impact.
How do we design, monitor and update our systems to reduce adverse impact?
Eliminating adverse impact and increasing workforce quality are not mutually exclusive. Select International's hiring systems are designed to balance the two goals simultaneously. In a review of multi-scale employee assessment tests, no predictor was found to have higher validity and lower adverse impact than Select Assessment for Manufacturing, one of our in-depth assessments that measures hourly level manufacturing candidates. To achieve this, we utilize multiple strategies and apply the following practices:
Using a combination of biographic measures, motivational fit data, applied problem solving tests, situational judgment and personality belief inventories. Measuring a full range of relevant competencies using a system consisting of various assessment tools can improve overall validity and minimize group differences.
Including interactive simulations and/or work samples that are face valid and realistic to the job. Research has shown that simulations reduce the majority and minority mean differences more than that of traditional paper-and-pencil tests.
Customizing assessment profiles with weights, minimums and cutoffs based on job analysis and validation study results. Candidates must possess a minimum level of aptitude in individual competencies and reach a certain level on the overall assessment score, calculated as a weighted average of the individual competencies.